Premium
DETONATION CLASSIFICATION FROM ACOUSTIC SIGNATURE WITH THE RESTRICTED BOLTZMANN MACHINE
Author(s) -
Bengio Yoshua,
Chapados Nicolas,
Delalleau Olivier,
Larochelle Hugo,
SaintMleux Xavier,
Hudon Christian,
Louradour Jérôme
Publication year - 2012
Publication title -
computational intelligence
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.353
H-Index - 52
eISSN - 1467-8640
pISSN - 0824-7935
DOI - 10.1111/j.1467-8640.2012.00419.x
Subject(s) - computer science , discriminative model , support vector machine , signature (topology) , artificial intelligence , task (project management) , preprocessor , pattern recognition (psychology) , machine learning , hyperparameter , relevance vector machine , speech recognition , data mining , engineering , mathematics , geometry , systems engineering
We compare the recently proposed Discriminative Restricted Boltzmann Machine (DRBM) to the classical Support Vector Machine (SVM) on a challenging classification task consisting in identifying weapon classes from audio signals. The three weapon classes considered in this work (mortar, rocket, and rocket‐propelled grenade), are difficult to reliably classify with standard techniques because they tend to have similar acoustic signatures. In addition, specificities of the data available in this study make it challenging to rigorously compare classifiers, and we address methodological issues arising from this situation. Experiments show good classification accuracy that could make these techniques suitable for fielding on autonomous devices. DRBMs appear to yield better accuracy than SVMs, and are less sensitive to the choice of signal preprocessing and model hyperparameters. This last property is especially appealing in such a task where the lack of data makes model validation difficult.